Syntactic flexibility refers to the fact that speakers can express most ideas with more than one syntactic structure. Such flexibility implies that the production system must choose which structure to use to express an idea, a process that is guided by two classes of factors. Information-processing factors, including accessibility (speakers use structures that mention accessible material earlier) and syntactic persistence (speakers use structures like those they have recently encountered), cause speakers to choose structures based on the cognitive or linguistic properties of the to-be-produced material. Communicative factors, including features of dialogue and discourse, cause speakers to select structures that facilitate information- transmission and coordination. These classes of factors are necessarily interrelated, as communicative influences must be exerted through the information-processing mechanisms that underlie syntactic production. The experiments proposed in Section I will use picture-naming and sentence-recall tasks situated in communicative settings to investigate the information-processing basis of sentence production by determining how accessibility and syntactic persistence effects are influenced by factors related to working memory, incrementality, lexical-syntactic interactions, and implicit learning. Specific theoretical questions that are addressed include how different working-memory constructs (storage, retrieval, executive functioning) influence sentence production, the scope of sentence planning, current accounts of the degree to which sentence planning is based on lexical versus lexical-pius-syntactic processing, and whether syntactic persistence reflects implicit learning or residual accessibilty. The experiments proposed in Section II will extend this approach to investigate how accessibility and syntactic-persistence effects are influenced by communicative factors including discourse and dialogue structure and disfluency. The broad theoretical question addressed in Section II concerns the degree to which the influence of communicative factors reduces to the influences of production-motivated information-processing factors, versus being fundamental cognitive primitives that influence sentence production. The proposed research promises important new theoretical and practical insights. The cognitive mechanisms to be explored in the experiments form the basis both of psycholinguistic theory and language development. Better understanding of how such mechanisms lead to typical language production and communication will constrain and develop existing theory, as well as guide the development of approaches to diagnosing language and more general disorders and improving treatment.